The study showed that (1) the OHS, and MCS-12 of the medial support group were significantly better than those of the non-medial support group; (2) the healing rate was relatively high and the healing time was shorter for the medial support group than for the non-medial support group; (3) the incidence of complications was lower for the medial support group, but no significant differences were observed for surgical trauma. Therefore, the reconstruction of medial support may be important for revisions of failed femoral trochanteric fracture treatments and might improve clinical prognosis.
At present, single intramedullary or extramedullary fixation are still used in many hip-preserving operations most commonly angle-stable plate, cephalomedullary nail, or locking compression plate (LCP) [1–2, 6, 19]. However, single implant for revision fixation is often unable to achieve a satisfactory clinical prognosis. Lambers et al.  revised 11 patients with failed trochanteric fracture treatment by replacing the cephalomedullary nail, resulting in a failure rate as high as 27.3%. Another commonly used fixator, dynamic hip screw (DHS), must be combined with a valgus osteotomy to achieve a relatively high healing rate. [14–15] However, this procedure alters the normal alignment of the lower extremities, increasing the risks of joint pain and degeneration and significantly increasing the difficulty of further salvage revisions . Lotzien et al [12, 21] did not use valgus osteotomy but applied DCS as the primary fixation method for treating patients following surgical failure of intertrochanteric fractures. The failure rate of revision surgery among the DCS group was 44.4% (4/9) .
In response to the above problems, the value of the medial support reconstruction of the proximal femur has gradually attracted surgeons’ attention. Finite element analysis and biomechanical studies have shown that the integrity of the medial wall is important for increasing the destructive load. Furthermore, compared with the reconstruction of the integrity of lateral wall in proximal femur, the integrity of the medial wall may be taken priority during the fracture fixation to achieve better stability. [22, 23]. Relevant clinical studies have pointed out that poor restoration of the medial wall of the femur will make the intramedullary nail fixation tend to fail . Xue et al.  also found the importance of restoring effective medial support at the intertrochanteric region in the treatment of intertrochanteric fracture nonunion.
Therefore, given the potential benefits associated with the reconstruction of medial support, MABP, as a specific implant that restores the medial support of the hip, has been designed and successfully applied to the treatment of femoral neck fracture and nonunion [25, 26]. Hence, to furtherly evaluate the importance of medial support reconstruction in hip-preserving revision surgery following failed treatment of inter/peri-trochanteric fractures, we conducted the present study. The clinical prognosis of patients in medial support group in our study showed higher functional scores, better fracture healing outcome, and less complications than ones in non-medial support group and the other previous related studies [11, 12, 21], which were contributed by better improvement of mechanical and biological environment on nonunion site simultaneously: (1) Direct medial support: Previous medial support plates are placed anterior to the proximal femur, only provide indirect support [11, 12]. While, MABP is placed just below the femoral neck and medial side of proximal femur to provide direct support. This placement position of MABP can better abolish the varus stress of the proximal femur, and provide more effective mechanical stability for nonunion site. This advantage was also preliminarily confirmed by biomechanical studies, which might benefit patients in terms of early functional exercises and fracture healing ; (2) Thorough debridement and autograft: The intertrochanteric region is rich in blood supply and cancellous bone. Hence, when revision is carried out, it is previously believed that thorough debridement and autograft of the nonunion site are not necessary . However, through our study, it was found that the success rate of fracture healing and one-stage revision surgery in both groups were significantly higher than previous clinical reports. This is most likely related to thoroughly debriding and autografting in all patients. However, it is worth noting that the failure rate of revision surgery remained 66.7% higher in the non-medial support group, and the insufficient improvement of mechanical stability carried by the single implant may contribute to this outcome.
However, this study also has some limitations. First, as a retrospective comparative study, this study is vulnerable to the inherent inclusion and exclusion biases that cannot be adjusted. Second, the relatively small patient sample size and improper control of confounding factors may affect the corresponding conclusions. Finally, a lack of consistency exists between implant species and fixation constructions. Although these differences in implant selection followed basic principles and standard surgical procedures were applied to treat nonunion, they remain associated with stability differences in the final fixation.